Volume 70, Issue 1, Pages 157-164 (July 2006) Laser Doppler flowmetry detection of endothelial dysfunction in end-stage renal disease patients: Correlation with cardiovascular risk  A. Kruger, J. Stewart, R. Sahityani, E. O'Riordan, C. Thompson, S. Adler, R. Garrick, P. Vallance, M.S. Goligorsky  Kidney International  Volume 70, Issue 1, Pages 157-164 (July 2006) DOI: 10.1038/sj.ki.5001511 Copyright © 2006 International Society of Nephrology Terms and Conditions

Figure 1 Comparison of risk assessment scores using the Framingham and Cardiorisk criteria (R2=0.575). Values at lower portion of risk spectrum were almost equivalent when calculated using either method. At the higher portion of risk spectrum, there was a tendency for Cardiorisk to produce a higher estimate than Framingham risk assessment. Results are shown as 10-year risk (%). Kidney International 2006 70, 157-164DOI: (10.1038/sj.ki.5001511) Copyright © 2006 International Society of Nephrology Terms and Conditions

Figure 2 CRP and homocysteine levels in ESRD patients with high and low CV risk. Risk was calculated using both (a and b) Cardiorisk and (c and d) Framingham and patients were assigned low- or high-risk category based on the binary conversion at the median. Results are shown as mean±s.e. Kidney International 2006 70, 157-164DOI: (10.1038/sj.ki.5001511) Copyright © 2006 International Society of Nephrology Terms and Conditions

Figure 3 Parameters of LDF and risk calculated using the Framingham criteria. (a–c) Postocclusion (PO) measurements were not significantly different between these groups. (d–f) High-risk patients (above the median) had prolonged times to reach the thermal peaks and the nadir although the differences were not significant. High-risk patients had significantly diminished amplitudes for the (g) first heat peak, (h) heat nadir, and (i) second heat peak. Kidney International 2006 70, 157-164DOI: (10.1038/sj.ki.5001511) Copyright © 2006 International Society of Nephrology Terms and Conditions

Figure 4 Laser Doppler flow data and risk calculated using Cardiorisk. (a–c) No significant differences were noted between the groups during postocclusion reactive hyperemia. (d–f) High-risk patients (above the median) had prolonged times to each thermal peak and the nadir, similar to the results seen with Framingham (Figure 3). (g–i) The amplitude of thermal hyperemic peaks and nadir were diminished in high-risk patients, but only the amplitude of nadir achieved significance. Kidney International 2006 70, 157-164DOI: (10.1038/sj.ki.5001511) Copyright © 2006 International Society of Nephrology Terms and Conditions

Figure 5 Thermal hyperemia and cardiovascular risk. The area under the curve (AUC) of TH compared with CV risk using (a) Framingham and (b) Cardiorisk. Using either criterion the higher-risk patients had markedly diminished area under the curve (AUC) (mean±s.e). Kidney International 2006 70, 157-164DOI: (10.1038/sj.ki.5001511) Copyright © 2006 International Society of Nephrology Terms and Conditions

Figure 6 Regression model of LDF and CV risk. Area under the TH curve for the first 30 min of heating is plotted vs 10-year risk (%) calculated using the Framingham criteria (R2=0.085). The results were analyzed using linear regression analysis. Kidney International 2006 70, 157-164DOI: (10.1038/sj.ki.5001511) Copyright © 2006 International Society of Nephrology Terms and Conditions

Figure 7 Predicting CV risk. ROC curves were plotted using mROC as described for both Cardiorisk and Framingham assessments of risk. (a) The area under the ROC curve for prediction of Framingham risk was 0.802 with a CI of 0.5–0.816. The area under the ROC curve for predicting (b) 10-year and (c) lifetime Cardiorisk are 0.802 with a CI of 0.5–0.810 and 0.758 with a CI of 0.5–0.747, respectively. The distribution of the results of the linear equation generated using all measured parameters is shown for low and high-risk patients as classified by (d) Framingham, (e) 10-year Cardiorisk, and (f) lifetime Cardiorisk. Kidney International 2006 70, 157-164DOI: (10.1038/sj.ki.5001511) Copyright © 2006 International Society of Nephrology Terms and Conditions

Figure 8 ESRD patients with abnormal LDF parameters and without known CV disease. The subset of ESRD patients with no known history of CAD or DM was grouped according to the results of their LDF study. The patients whose area under the TH curve was lower than the median were grouped as ‘low perfusion’. Those patients above the median were termed ‘high perfusion’. Low perfusion patients had somewhat increased risk assessments using both the (a) Cardiorisk and (b) Framingham criteria, although the differences were not significant. No statistically significant differences in (c) CRP levels or (b) homocysteine levels existed between low- and high-perfusion groups. Kidney International 2006 70, 157-164DOI: (10.1038/sj.ki.5001511) Copyright © 2006 International Society of Nephrology Terms and Conditions

Figure 9 Risk assessment in ESRD patients with and without known CV disease. Patients were grouped according to their history of known CAD and or DM; these groups were then compared with (a) Cardiorisk and Framingham risk assessment, (b) CRP and homocysteine levels', and (c) LDF parameters. Varying levels of statistical significance were achieved for these comparisons, as indicated. Certain values (CRP, postocclusion area, time to thermal peak 2, and the area under the TH curve) were modified as indicated to fit the ordinate. HC=homocysteine, PO=Post Occlusion, Amp=amplitude, HP1/HP2=first and second heat peaks. Kidney International 2006 70, 157-164DOI: (10.1038/sj.ki.5001511) Copyright © 2006 International Society of Nephrology Terms and Conditions

Figure 10 CV mortality was associated with a significant suppression of the amplitude of the first and second thermal hyperemic peaks. Kidney International 2006 70, 157-164DOI: (10.1038/sj.ki.5001511) Copyright © 2006 International Society of Nephrology Terms and Conditions

Figure 11 Laser Doppler parameters and their changes in ESRD patients. Laser Doppler flow averages of control and ESRD patients are shown for PORH and thermal hyperemia experiments (±s.d.). The ESRD average perfusion (green line) is well below that for the control population (red line) throughout both portions of the experiment. PORH measurements included (a) amplitude and (b) time to peak as well as the (c) area under the curve. Thermal hyperemia measurements included (d–f) amplitudes and (g–i) times to first heat peak, heat nadir and second heat peak, respectively. Additionally, the area under the curve for the first 30 min of heating was measured. Kidney International 2006 70, 157-164DOI: (10.1038/sj.ki.5001511) Copyright © 2006 International Society of Nephrology Terms and Conditions